Processing Dynamic Event Data and Multifaceted Knowledge in a Collaboration Federation

University Of Florida, Gainesville FL

Investigators

Abstract

All nations are facing global problems such as border control and immigration, drug trafficking, terrorism, human disease detection and control, rapid detection of diseases and pests threatening agricultural commodities, and others. The solutions to these complex problems require organizations within a country as well as across national boundaries to effectively and efficiently share, not only heterogeneous data, but also knowledge captured in different knowledge representation schemes and application systems to support their problem-solving and decision-making in highly distributed and dynamic environments. This project will develop the infrastructure and technologies for the processing and management of 1) distributed events, 2) dynamic data associated with event occurrences, and 3) knowledge expressed by different types of rules and rule structures (multi-faceted knowledge) and application system operations invoked by rules. The context for this research is the USDA's National Plant Diagnostics Network (NPDN), a network established for rapid detection of crop disease and pest outbreaks. NPDN's goal is to enable many collaborating organizations to receive data associated with each occurrence of an event type as well as data generated by different types of rules and rule structures, and application operations triggered by the event. This will support the local decision-making, problem-solving and activity coordination of collaborating organizations. The expected results include 1) an extended rule markup language and user interface tools for defining events and capturing multi-faceted knowledge, 2) techniques for efficient and effective management and processing of distributed, heterogeneous events, event data, triggers and rules, 3) a domain ontology for plant disease diagnostics and techniques for ontology management, 4) an extended Web Service infrastructure with an ontology-enhanced and constraint-based registry for semantic discovery of triggers, rules and application operations that are uniformly modeled and published as Web Services, and 5) the deployment of the developed tools and system at several regional centers of the NPDN for application and evaluation of the R&D results. Intellectual merits: First, the extended Rule Markup Language will contribute to the effort of the RuleML community in developing a standard language for organizations to specify and share policies, regulations, processes and constraints expressed in terms of high-level declarative rules. Second, the resulting techniques, algorithms, methodologies and infrastructure will significantly advance the state of arts of information and knowledge management. Third, the proposed ontology-enhanced and constraint-based Web Service registry will significantly improve the quality of discovery and invocation of registered rules, rule structures, and application system operations. By converting heterogeneous rules and rule structures into Web Services, all that will be needed is a single rule server that invokes them as Web Services to achieve collaborative problem-solving. Broader Impacts: First, the R&D results will be applied and evaluated by a federation of collaborating organizations in the NPDN environment. The deployment of the resulting collaborative system and tools will have an immediate application to and impact on the NPDN organization because they will be timely informed of any disease or pest outbreak and receive guided assistance on appropriate emergency response resulting from the application of knowledge rules. Second, events, multifaceted knowledge specifications and application operations registered in the Web Service registry will capture collaborating organizations' policies, regulations, processes and constraints for later education and training uses.